Maltreated Children’s Memory for a Traumatic Separation:
Relations between Stress, Dissociation and Memory for the Event
Ragnhild Klingenberg Stokke
Department of Psychology University of Oslo
Norway
2007
Levert som hovedoppgave ved Psykologisk Institutt Universitetet i Oslo, 2007
Abstract Author: Ragnhild Klingenberg Stokke
Title: Maltreated Children’s Memory for a Traumatic Separation: Relations between Stress, Dissociation and Memory for the Event
Supervisor: Associate professor Annika Melinder
This thesis serves as a pre-study of the research project ”Children’s Memory for Traumatic Separations: An investigation of children removed from home by the Child Protective Services”. The author has contributed in all steps of the study, such as in the original planning, in data collection, and in training, coding, and reliability testing of all dependent measures employed. The author has conducted the child interviews and testing.
There is an ongoing debate whether memory for traumatic events can be lost, or if there are unique influences such as dissociation (e.g., the disruption of normal integration of memories, perceptions, and identity into a coherent sense of self) superseding general memory mechanisms (e.g., age differences, forgetting, and memory illusion). The impact of maltreatment-related sequel on basic memory processes is unsettled. Previous studies of trauma and memory have primarily been field research projects. Very few, if any, reports exists that both hold a high ecological validity and employ experimental demands, which enables full control of the situation to be studied. The present thesis is therefore a pioneer work, as it investigates real life phenomenon employing an experimental design.
To study memory for real life traumatic events within a cognitive developmental approach, a removal situation was chosen as the event to be encoded and later recalled, and 12 maltreated children aged 3 to 12 years old were recruited. At the day of removal a researcher observed and registered the child’s and the parents’ reactions, and the placement procedure.
The children accomplished a structured memory interview one week and three months after the removal day, Child Behavioral Checklist and Trauma Symptom Checklist for Young Children were filled out, and cognitive tests were taken. Biological parents and the CPS caseworker were interviewed, and case report information was registered.
Results showed that degree of stress experienced during removal related to accuracy in the children’s memory. Mixed results were found regarding age and the amount and accuracy of information given, and between memory and PTS symptoms. Due to dissociation, a
slightly negative impact on memory was found. Preliminary results are discussed in light of previous research on maltreated children’s memory for trauma and corresponding theories.
Preface
Since the beginning of this project in the spring of 2004 I have learned a lot about planning, preparing, and executing a research project; research being a time consuming process
necessary to experience and give time to mature. During the course of the project I have spent countless hours with my supervisor and fellow researcher discussing research perspectives and ethical dilemmas during the original planning and as it processed; designing measures for data collection; getting the necessary public permissions from the Regional Committee for Medical Research Ethics, the Ministry of Children and Equality, the Board of Confidentiality and Research, and the Data Inspectorate; presenting the project for the cooperating Child Protective Services; collecting the data; adjusting existing coding schemes to our interview material, training to be reliable in its use; coding and scoring the material; conducting statistical analyses; presenting the project and results at national and international conferences; and the writing process. Being a part of this research project have made it possible for me to combine two main areas of interest, maltreated children and the Child Protective System in Norway, and cognitive and psychopathological developmental psychology.
In this process several people have provided valuable support and professional advice.
My thanks goes to Professor Gail Goodman for her advice on planning the study and developing the interview guide for data collection; PhD student Gunn Astrid Baugerud for valuable discussions, great cooperation, and transmitting energy to proceed this research project; colleagues in the Child Protective Services for useful feedback; members of the EKUP lab group for valuable help and thoughtful comments; and my brothers for encouraging support.
My greatest thanks go to my mentor and supervisor Annika Melinder for her
dedication, support and inspiration throughout these years. A special thank for her courage to develop a challenging research project in light of the importance to gain more knowledge about an especially vulnerable group in our society, maltreated and traumatized children.
During these years Melinder has also given me the opportunity to develop my academic skills through participation in courses, lab groups and conferences, which I am grateful for.
This project was supported by grants from the Ministry of Children and Equality to the Cognitive Developmental Research Unit led by Annika Melinder, and Norwegian Research Council student research fellowships to me.
Table of Contents
Abstract II
Preface III
Table of Contents IV
Introduction 1
Memory systems – autobiographical memory 2
Development of autobiographical memory in childhood 3 Theories of autobiographical memory in childhood 4
What characterizes a traumatic event? 5
Current theories of trauma memory, and dissociation 6 Trauma and memory in preschool and school-aged children 11
The concept of dissociation 15
Child maltreatment, memory, and dissociation 17
Present study 19
Method 20
Ethical considerations and informed consents 20
Design 22
Participants 22
Measures 23
Procedure 28
Confidentiality 29
Results 30
Discussion 36
Effect of age on memory amount and accuracy 36
The relation between stress and memory amount 38
The relation between stress and memory accuracy 39
Dissociation and memory for the separation 40
Associations between short-term memory and dissociation 43
Limitations and challenges 43
Preliminary conclusions 46
References 48
Appendix A Sequential schema for observation Appendix B Child Memory Interview 1
Introduction
“I know that Lisa and Torgeir from the Child Protective Services picked you up and you moved here. I wasn’t there, so I would like you to tell me everything that you remember from the day that they came and you moved here.” Miriam, 6 years answers: “The police came to our house, where I used to live.
Mummy cried. I didn’t want to move. Mummy was looking, my sister playing computer games. They just picked me up. They said I ought to come to this place (a stand-by home). Then we sat in the car.
The CPS took my clothes and brought it here (silence). We should move.”
The researcher observed the girl having strong reactions when the CPS explained to her what was going to happen. She seemed frightened and said: ”But I can’t move, I can’t move”, ”mummy will not allow me to move”, and ” I’m not allowed to speak to you in the CPS”. The girl was difficult for the CPS workers to get in touch with, seeming to be in a daze, her eyes were kind of disappearing, she closed them and it was like she fainted.
Traumatic events, such as the experience this child describes, raise important questions about memory for trauma at different levels; at a societal level (e.g., witness testimony, mental health perspective); at an individual level (e.g., the experience of memory gaps, and flashbacks), and; at a theoretical level (e.g., what influences the encoding and retrieval of stressful and traumatic events may have on memory). Every day, children experience being removed from their parents by the Child Protective Services (CPS). It is important to gain knowledge about how the procedures for these removals may be improved, how the individual child’s memory for the event is shaped and further developed, and whether or not there are unique influences on memories for traumatic events such as dissociation. Will Miriam in the example above be able to recall this day the following months? What will be remembered for later recall, and will the child’s memories get influenced if symptoms of dissociation are present during or shortly after removal? It has been a huge debate about whether traumatic experiences in childhood can be repressed or lost, or whether special memory mechanisms such as dissociation supersede normal memory processes in a way that make memories of the event inaccessible for an extended period of time (Goodman et al., 2003). This thesis has a cognitive developmental approach using an experimental design to study the trauma of separation and replacement of preschool and school-aged children. It is a pre-study in an extensive longitudinal project. As far as known,
experimental studies of a real life separation has never been conducted on this group of children before.
In the following sections, the focus will be on how children experience and remember being removed from home by the Norwegian CPS and placed in a suitable care facility. These children have a history of maltreatment and/or severe abuse in addition to the experience of being separated from their biological parents. Children with a history of trauma and maltreatment (e.g., psychological abuse, physical abuse, sexual abuse, neglect, and substance abuse) represent an especially vulnerable group. It is important to gain more knowledge about how this group perceives being removed from home, including if their memory for the situation is impaired, enhanced, or if their memory for the traumatic
separation may predict later psychological functioning. Trauma and the stress associated with it are thought to have a variety of effects on memory for traumatic events themselves as well as for basic memory processes (Howe, Toth, & Cicchetti, 2006b). Children are thought to be especially prone entering temporary dissociative states to cope with intense stress or trauma (Putnam, 1997).
Dissociation, the disruptions of normal integration of memories, perception, and identity into a coherent sense of self, has profound implications for young children’s memory for traumatic experiences and is therefore of particular interest for trauma researchers, cognitive scientists and clinicians (Cordón, Pipe, Sayfan, Melinder, & Goodman, 2004; Macfie, Cicchetti, & Toth, 2001). Before discussing factors influencing memory for traumatic events and if, and how, dissociative processes may affect such memories, a description of the memory system, particularly autobiographical memory, and theories on trauma memory including dissociative processes, is provided.
Memory systems – autobiographical memory
Memory theories include two main distinctions, characterizing memory by retention time giving us sensory, short-term and long-term memory, or by their content giving that different types of information can be retained in partially or wholly distinct memory systems departing (Gazzaniga, Ivry, & Mangun, 2002). Long-term memory systems are usually described as divided in explicit (declarative) memory referring to knowledge we have conscious access to, and implicit (non-declarative) memory referring to procedural knowledge such as motor and cognitive skills. Explicit memory can further be divided into episodic and semantic memory, a distinction introduced by Endel Tulving (1972). Memory for events in one’s personal past occurring in a specific time and place having sensory recollections associated make up the autobiographical (episodic) memory (Nelson & Fivush, 2004). According to Tulving (1985), episodic memory depends on the ability to mentally travel back in time, and with the term “autonoetic”
Tulving referred to the special kind of consciousness that allows humans to be aware of subjective time when an event took place. In contrast, semantic memories reflect the person’s world knowledge such as language and facts (Gazzaniga et al., 2002). The distinction between episodic and semantic memory, has been supported by research using fMRI technology observing increased activity on both sides of the frontal lobes when using the episodic memory, in contrast to a single side activity when performing semantic memory tasks. These findings support the existence of two different neurological systems in long-term memory explicit division (Tulving, 2002). However, autobiographical and semantic memories draw on each other e.g., making an assimilation of the autobiographical memory according to the
semantic knowledge of the world. In this way, experiences are linked with earlier knowledge and both subsystems might be adjusted (Nelson & Fivush, 2004).
The process of memory consists of three major hypothetical stages: encoding, storage and
retaining. Encoding refers to the processing of incoming information to be stored, happening through the registration of input in sensory buffers and sensory analysis stages (acquisition) and the creation of a stronger representation over time (consolidation) resulting in storage of the memory (Gazzaniga et al., 2002). Encoding is affected by multiple factors, such as the child’s earlier experiences and prior knowledge of the world (Cordón, 2002, cited in Cordón et al., 2004), which, in turn, affects how an event is interpreted and understood. During early childhood extensive developmental achievements are reached, thus expected to affect the memory reports given. What information being encoded depends on cognitive processes like attention directed by selection and concentration within the limit of its capacity (Gazzaniga et al., 2002). For memories to be explicitly available for recall later, they must be integrated and consolidated (Bauer, 2007).
Development of autobiographical memory in childhood
Events experienced before the age of about 18 months are found not to be accessible verbally, while for the age span 18 months to 2.5 – 3 years children are able to provide coherent reports of events but in a brief, fragmentary fashion and prone to increasing error over time (Bauer, 2004; Fivush, 1998).
Experiences that are not available for explicit recall in childhood are seldom found to be a part of adult autobiographical memory (Cordón et al., 2004). Concerning what is known as infantile amnesia, theories are many. One suggests that memories are formed before age 2, but later become inaccessible as a result of cognitive changes, e.g., the onset of language (Bauer, 2004). Other theories point out the development of the “cognitive self” enabling children from around 2 years of age to process events like something that happened to “me” (Howe, Courage, & Edison, 2003), and starting to form
autobiographical memories. Thus, the inaccessibility of early memories, traumatic or otherwise, is suggestively explained in terms of cognitive, neurological, linguistic, and social factors (Bauer, 2007;
Cordón et al., 2004; Howe et al., 2006b).
By about 3 years, children start talking about past events more independently from adult scaffolding, and begin to use the story or narrative form in these conversational interactions (Howe et al., 2003). Individual differences exist, but from this age and with gradually more sophisticated
language, memories can be retained and organized around a life history including concepts of time and place (Fivush, 1998; Howe et al., 2003). Research studies have included children in an age range enfolding infants and early childhood, e.g., Howe et al. (1994). Children ranging in age from 18 months to 5 years were interviewed about their memories for emergency room experiences following injuries such as fractures, lacerations, and severe burns. The interviews where conducted a few days after the event and again 6 months later. Children younger than 30 months at the time of injury and hospital visit recalled little at either interview, whereas the children older than 30 months were able to report their
experiences at both interviews (Cordón et al, 2004; Howe et al., 1994). This illustrates when the ability to verbally recall a personal relevant event occurs.
Memory develops according to other aspects of neurological, cognitive structures and linguistic ability, making memory more effective with age given the child’s increase in world knowledge and consciousness about personal mental processes. For example, children’s ability to code sources of information, the achievement of better language, and a repertoire of acquisition and organization
strategies (Howe, Cicchetti, & Toth, 2006a). The principles of memory development are mainly derived from research on non-traumatized children recalling pleasant events like visits to a museum or shopping mall, or trips to amusement parks (Howe, Cicchetti, Toth, & Cerrito, 2004), but complementary research on traumatized children has increased, especially regarding childhood sexual abuse, giving raise to theory development on childhood memory. There has been considerable debate about when episodic memory is first available (Goodman & Melinder, 2007a) and how it emerges through childhood.
Theories of autobiographical memory in childhood
Different models of autobiographical memory development exist (Goodman & Melinder, 2007a; Nelson & Fivush, 2004). Howe and colleagues suggested a model focusing the development of a sense of self, which they term ”the cognitive self” (Howe et al., 1994). The cognitive self appears around the age of 2 years, making the child able to organize
information and experiences as something personal happening to ”me”, recognizing it self as part of the event. This is suggested to contribute to the gradual ending of infantile amnesia as the child starts making a cognitive self-schema (Goodman & Melinder, 2007a). Disturbance in the development may happen, as that of delayed maturation (e.g., due to Down syndrome, familial mental retardation). These children acquire a cognitive self if, and when, they achieve a mental age comparable to that of non-delayed infants (around the age of 2 years) (Howe et al., 2006b). Not found to be affected by child maltreatment, the onset of cognitive self seems to be linked to maturation-constitutional factors in a higher degree than social-experiential factors (Howe et al., 2006b). Other theorists agree that the development of ”a sense of self” is an important factor in the development of verbally accessible memories, not conflicting with the view of sociolinguistics adding an emphasize on language and narrative skills as well (Goodman & Melinder, 2007a; Nelson & Fivush, 2004).
Nelson and Fivush (2004) point out that these components operate within cultural and social contexts, where maternal elaboration of the child’s autobiographical experiences together with the child, called adult memory talk, is playing a particularly important role.
Typically, the parents start to talk to their child about everyday life events as the child
develops a sense of self, coincidental with the child’s language comprehension and expression development exploding in the middle of the second year. This developmental stage is seen as especially important as it fosters sequential thinking and temporal organization (Nelson &
Fivush, 2004). Differencing from Howe et al. (2006b) introducing the development of ”the cognitive self” as the break point of autobiographical memories, Nelson and Fivush (2004) promote a slowly development of the formation toward an adult autobiographical memory from the general beginning around 3 or 4 years of age, with only a few memories from each year until school age. In addition, sociolinguists emphasize the child-parent relationship in the maternal reminiscence style, but without concern of the value of this relation.
Goodman and Melinder (2007a) suggest, in light of the highly agreed upon
importance of sense of self in the development of autobiographical memory and the role of maternal elaboration, that both the parent’s and the child’s attachment orientation influence the development of autobiographical memory. The model postulates that attachment relations play a central role in the cognitive and verbal processing of events in the childhood, especially negative experiences as it is thought to elicit attachment behavior for survival reasons.
Children’s own attachment orientation linked to affect regulatory processes, and parental attachment style affecting if and how parents help children to talk and think about the negative experiences, boosting the processing (Goodman & Melinder, 2007a). The model further proposes that enduring autobiographical memories are likely acquired between the 2nd and the 4th year of age as a function of brain maturation, influenced by socio-emotional and cognitive environment (Goodman & Melinder, 2007a).
Relating the above models and debates of autobiographical memory to the present study, further writing will be concerned with memory development once it is possible to form an enduring autobiographical memory. Young children (3 to 6 years of age) are expected to report less information about the trauma according to both the theories above.
Before describing current theories and knowledge in the field of memory for traumatic events in maltreated and non-maltreated children, a definition of trauma is demanded.
What characterizes a traumatic event?
What constitutes a traumatic experience is lacking a clear definition in the literature existing.
According to DSM-IV-TR, trauma involves witnessing, experiencing or being confronted with "actual or threatened death or serious injury, or a threat to the physical integrity of oneself or others" (American Psychiatric Association (APA), 2000, p. 463). The exposure leads to a response involving intense fear, helplessness, or horror, which in children can be
expressed by disorganized or agitated behavior (APA, 2000). Trauma has been variously described, e.g., as an experience that: (1) threatens the health or wellbeing of an individual (Brewin, Dalgleish, & Joseph, 1996); (2) indicates that the world is an uncontrollable and unpredictable place (Foa, Zinbarg, & Rothbaum, 1992); and, is an inescapably stressful event that overwhelms an individual’s coping mechanisms (van der Kolk & Fisler, 1995). The child’s world knowledge and earlier experiences affect how the child understands and
interprets events, mundane as well as traumatic, contributing extensively to whether the child perceive and experience the event as traumatic (Cordón et al., 2004; Pipe & Salmon, 2001). In contrast to the ICD-10 defining traumatic experiences in objective term, the stressor definition in DSM-IV-TR includes a subjective component including descriptions of suspected
responses to the trauma (APA, 2000; Diseth, 2005; WHO, 1992). This component acknowledges that the personal reaction plays a crucial role (Salmon & Bryant, 2002).
Another perspective taken to conceptualize trauma, is to what degree these events (such as child abuse, sexual assault) involve social betrayals. A growing body of research
demonstrates that events high in betrayal are associated with significant distress, as would be expected if these events were traumatic by the more common use of the term (Freyd,
DePrince, & Gleaves, 2007).
Which situations children experience as traumatic compared to stressful or painful, differs across subjects, but a forced removal from home may be a traumatic experience (Leslie et al., 2000). The separation from their family and loss of their parents may contribute to undermine the children’s sense of belonging and even further impair their presumed already battered sense of self-esteem (Davidson-Arad, Englechin-Segal, & Wozner, 2003). A feeling of fear and helplessness can be expected in children being removed from home, like Miriam saying she does not want to move, knowing that the CPS decides independently of what she or her mother say. Miriam cannot escape the situation. Many removals also happen acute, and for planned ones the child is not always informed or just partly informed about the removal, giving an unpredictable aspect to the situation.
Current theories of trauma memory, and dissociation
As one of the early theorists of dissociation in the present, Putnam (1997) proposed that dissociative behavior and trauma-related symptoms are associated to the severity and persistence of the trauma. Psychological traumas occurring early in life, affects the child’s opportunity to follow a normal developmental trajectory affecting basic conditions like the attachment system (Nijenhuis, van der Hart, & Steele, 2006). Theories concerning
dissociation differ in their perspective, e.g., whether dissociation is the mechanism behind recall failure, or if dissociation provides a potential explanation of memory phenomena linked to PTSD or autobiographical memory in general. For the purpose of the present thesis, three theories representing the width of theoretical thinking on this field are presented.
Betrayal trauma theory (BTT). Taking a survival perspective on children experiencing traumas from their caregivers, Freyd developed betrayal trauma theory positing that children remain unaware of caregiver-perpetrated abuse because this allows them to maintain the vital attachment to their caregivers (Freyd, 1996). Betrayal trauma occurs when the people or institutions on which a child (or adult) depends for survival violate that person in a significant way. Examples of betrayal trauma are childhood physical, emotional, and sexual abuse
perpetrated by a caregiver. Children who grow up in abusive homes develop divided attention skills that help them keep threatening information out of awareness, that is, they develop the means to ignore the abuse (i.e., dissociative abilities) (Becker-Blease, Freyd, & Pears, 2004b).
Specifically, the theory proposes that the way in which events are processed and remembered will be related to the degree to which a negative event represents a betrayal by a trusted, needed other (Freyd et al., 2007). Theorists within this approach have proposed that memory impairment for trauma-related information involves avoidant processing, e.g., people may fail to encode the material (vs. impaired retrieval processes, e.g., McNally et al., 2005). DePrince and Freyd (2004) tested college students, some reporting childhood sexual abuse (CSA), under divided attention conditions. They found that students scoring high on a dissociation questionnaire exhibited memory deficits for trauma words (e.g., incest) when these were viewed under divided attention conditions. In addition, high dissociators reported significantly more trauma history and betrayal trauma (i.e., abuse by caregiver) than students low on
dissociation. The results support BTT predicting that adult survivors of sexual abuse who were molested by their caretakers are especially likely to dissociate their memories of abuse.
However, traumatic events tend to be remembered and may even be better
remembered (Cordón et al., 2004; Fivush, McDermott Sales, Goldberg, Bahrick, & Parker, 2004). One study replicating DePrince and Freyd’s procedure did not find support for the betrayal trauma theory. McNally and colleagues (2005) utilized the divided attention paradigm in testing for memory deficits for trauma words relative to neutral words in adults reporting either continuous or recovered memories of CSA versus adults denying a history of CSA (McNally, Ristuccia, & Perlman, 2005). Memory deficit for trauma words under divided attention was expected in the recovered-memory group, but the results were found to be inconsistent with this prediction, as all three groups exhibited better recall of trauma words
than neutral words, irrespective of encoding conditions (McNally et al., 2005). There is an ongoing debate between the two approaches. Goodman and colleagues (2003) found no relationship between abuser status (parent/caregiver vs. stranger) and failure to report abuse years after the abuse in a sample of 175 young adults who had participated in criminal proceedings related to sexual abuse allegations approximately 10 years earlier. Thus, as for most, if not all, survivors of trauma, little support has been found for the notion that trauma experiences are not encoded resulting in amnesia (Howe et al., 2006b). In the current study, no measures or tasks according to divided attention was included.
The next theory described makes a distinction between explicit (verbal) and implicit (non-verbal) memories, consistent with distinctions in basic memory system theory
(Gazzaniga et al., 2002), and suggests that there is no encoding of explicit memories during a trauma, however implicit memories are encoded and preserved.
Van der Kolk’s theory. The starting point for this theory is that traumatizing occurs when the individual’s internal and external resources are both inadequate to cope with an external threat.
According to van der Kolk (van der Kolk, 1994; van der Kolk & Fisler, 1995) mental traces after traumatic experiences are qualitatively different than memories of mundane events. The suggestion is that trauma interferes with explicit but not with implicit memory because trauma leads to the release of stress hormones, hormones that create a sort of state-dependent memory for the traumatic experience (van der Kolk & Fisler, 1995; van der Kolk, Hopper, & Osterman, 2001). These state-dependent memories are, according to van der Kolk, inaccessible to conscious recollection until the same state is induced again and the traumatized start to talk about the sensations, remaining intact and unchanged until that happens (Howe et al., 2006b; van der Kolk et al., 2001). Research conducted by van der Kolk and Fisler (1995) during provocations of traumatic memories, found that a lowered activity in the language area during re-experience of trauma causes the failure of encoding explicit memories of the event, leaving more primitive organizations of the experience e.g., visual pictures and somatic sensations (van der Kolk, 1994; van der Kolk & Fisler, 1995). In another study van der Kolk and colleagues have found that traumatic memories, especially those associated with PTSD, initially lack narrative elements. This was found in a study of 16 adults with and without current PTSD who had experienced awakening from general anesthesia during surgery (van der Kolk et al., 2001). Using the Traumatic Memory Inventory to assess the way memory for traumatic events are organized and retrieved at three time-delays, they found that participants with PTSD were more likely to report that they did not have a narrative memory initially, and had a higher degree of reliving of sensations and affects (van der Kolk et al., 2001). Observations from the clinic experiencing traumatized patients having trouble verbalizing their feelings support this theory (van der Kolk et al., 2001). Unlike explicit
memories, implicit memories are said to appear spontaneously in the guise of dreams, flashbacks, body sensations, avoidant behavior, and so forth (Nijenhuis et al., 2006). Thus, van der Kolk postulates that traumatic events initially are encoded as sensations or feeling states, and later once people start to talk about these sensations trying to make meaning of them or relive them, the memories are transcribed into narratives (explicit memories) (van der Kolk et al., 2001).
Research have found that stress actually aids the consolidation of memories for traumatic events and persist more accurately than for mundane events (see e.g., Peterson &
Whalen, 2001). One study found that higher degree of stress during trauma was associated with better memory and giving more information about the traumatic incident. This was in a study of children’s long-term memory for the hurricane Andrew (Fivush et al., 2004). Fivush and colleagues interviewed 3-4 year old children experiencing the natural disaster within a few months after the trauma, and again 6 years later when they were 9-10 years old. Children were grouped according to level of severity of the experience giving high, moderate or low degree of stress (e.g., at home when their family’s house fell apart, experiencing flooded basements and trees knocked down, or no storm, just heavy rainfall). Findings showed that all children reported the event vividly at both delays (Fivush et al., 2004). Howe et al. (2006b) agree with van der Kolk that extreme levels of stress can impair consolidation of an event, but comment that even if trauma resulted in state-dependent memory that could not be
consciously accessed, evidence has not supported that individuals become amnesic for the experience. Individuals can report dissociative alterations in consciousness like slowing of time and “out of body”-experiences, but can still provide declarative recollections (Howe et al., 2006b). Van der Kolk’s suggestion that implicit memories remain intact until the same state is induced has not been supported by research so far (Howe et al., 2006b).
Implicit, like explicit, memories are found to be subject to change and distortion. Not many studies have included implicit memory for traumas in their assessment though, but in a study conducted by Goodman and colleagues (1997), children were first asked to recall verbally their experience from the painful medical procedure VCUG and then re-enact it with dolls and props (e.g., an anatomical doll, catheter tube). In this study, reporting of the main event was greatly facilitated by all age groups (Goodman, Quas, Batterman-Faunce,
Riddlesburger, & Kuhn, 1997).
Van der Kolk’s theory is similar to the Betrayal Trauma Theory (BTT) in that it suggests a total failure to encode traumatic experiences verbally, but as BTT suggests that the failure involves memory per se, van der Kolk’s theory postulates that traumatic memories are encoded implicitly. Another perspective on trauma and memory is the network theories
suggesting that traumatic memories are encoded but in somewhat different ways.
Dual representation theory (DRT). According to Brewin (e.g., 2003), memory consists of two representational systems giving verbally accessible memories and situationally accessible memories.
Before this multi-representational theory, Foa and colleagues were early in their suggestion of a fear network model (see e.g., Foa & Kuzak, 1986). Trying to explain findings like the notion that
traumatized individuals develop a heightened sensitivity to and a heightened memory for trauma-related information (see e.g., Howe et al., 2006b), they suggested that traumatized individuals develop semantic
“fear” networks that serve to organize trauma-related information. The network serves to preserve information about trauma (e.g., through rehearsal of information) and link similar experiences in memory, thus, making stronger traces (Foa & Kuzak; 1986; Reisberg, 2001). Research supporting aspects of this theory has been conducted. Individuals with documented CSA and individuals with more PTSD symptomatology have been found to have particularly accurate memories of the abuse (Alexander et al., 2005). The fear network model focuses on a single explicit format of mental representations, treating verbally accessible and non-verbally accessible information in the same way. Brewin has suggested a multirepresentational theory.
According to Brewin (e.g., 2003), the first of the two types of representations in the dual representational theory (DRT) reflects the individual’s conscious experience of the traumatic event, called verbally accessible memories (VAMs). The second type of
representation consists of situationally accessible memories (SAMs). VAM representations are fully contextualized within the person’s autobiographical database including sense of present and past, while SAM representations are not. SAMs are characterized by reliving in present, fragmented sensory video “clips”, and are not in context. DRT further suggests that VAM and SAM representations are encoded in parallel at the time of the trauma and between them if repeatedly exposed (Brewin, 2003). For example, holistic, dissociative memories or flashbacks, dreams, and trauma-specific emotions would be considered to be the result of the activation of SAM representations (via cueing), whereas people’s ability to recount the trauma, for example in recollection tasks for research purposes, their answers and narratives would be a function of accessing VAM representations (Dalgleish, 2004). Every incidence of re-experience in a normal recovery process leads to some information from the SAM being copied to the VAM, making them less overwhelming to the SAM system, giving that the individual get more control over the traumatic event (Brewin, 2003). A failure of the slow but consistent copying process through e.g., deliberate avoidance, can lead traumatic memories to remain in the SAM system giving symptoms of PTSD (Brewin, 2003). Dissociation is in this theory seen as a risk factor to develop PTSD.
Empirical evidence found by the same group claimed that two types of trauma memory could be detected in parallel in the same individual. In a series of experiments designed to test the theory Holmes, Brewin, and Hennessy (2004) had participants watch a trauma film under different conditions. In one condition, they had to carry out a concurrent visuospatial task, tapping a pattern on a concealed keyboard. In another condition, they had to carry out a concurrent verbal task, counting backwards in threes. The participants then had to record in a diary the number of intrusive memories of the film they experienced over the next week. The prediction was that the visuospatial task would compete for the resources of the SAM system, leading to perceptual information being less well encoded and resulting in fewer intrusions than a no-task control condition. In contrast, the verbal task was expected to compete for the resources of the VAM system, leading to a less-detailed conscious representation and
resulting in more intrusions than in the control condition. As in an earlier study (Brewin &
Saunders, 2001), the concurrent visuospatial task reduced intrusive memories the following week, but Holmes et al. (2004) also showed that the verbal task increased the number of intrusions relative to a control condition. These results support the claim of DRT.
The dual representational theory has been found to be useful both as a theoretical tool for generating research and for presenting a framework for therapeutic interventions for PTSD (Dalgleish, 2004). Similar to van der Kolk’s theory, the DRT build on two memory systems, but unlike van der Kolk’s theory suggesting that non-accessible memories remains intact while verbal traces fail to encode, the DRT suggests a parallel encoding in situationally and verbally accessible memories. A further difference is that Brewin postulates that exposure leads to a stepwise exchange of information about the trauma moving from the SAM to the VAM making the traumatic memory gradually more accessible and contextualized.
One limitation in van der Kolk’s and Brewin’s theories is the lack of a developmental perspective. Freyd and colleagues focus on childhood traumas, but much of the research conducted on childhood trauma do not support the notion that traumatic memories like the one studied in this pre-study is not encoded at all. The main focus is for treatment purposes (Dalgleish, 2004), which is of major importance. However, the different models lead to different predictions according to children and their memory for traumatic events and will be used to shed light on different possibilities when discussing the results of this pre-study.
Trauma and memory in preschool and school-aged children
Are memories for distinctive, traumatic events in some way different from memories of non- traumatic, mundane events? Do the variables found to influence children’s memories of
neutral or positive experiences influence memories of traumatic events as well, or are unique mechanisms affecting trauma memories? Over the last decades, extensive research have been done trying to find what factors influence children’s memory for traumatic events like natural disasters, medical procedures, accidents, and injuries (see e.g., Cordón et al., 2004; Howe et al., 2006a), studies of adults and their memory for child abuse (e.g., Alexander et al., 2005;
Goodman et al., 2003), and experimental studies of memory in non-maltreated and maltreated children (e.g., Becker-Blease et al., 2004b; Eisen, Qin, Goodman, & Davis, 2002). An
individual factor found to affect several aspects of children’s memory, is age.
Age. After the age of three, age at the time memory is encoded and assessed is found to be a reliable predictor of memory for mundane as well as stressful and traumatic events, particularly when children are asked for open-ended or narrative accounts (see e.g., Cordón et al., 2004; Fivush, 1998). Age effects are found related to the accuracy of memory reports and suggestibility, with younger preschool children doing more errors than older children (Eisen et al., 2002; Goodman et al., 1994; Quas et al., 1999). In one study, 189 children aged 3 to 17 years of age involved in evaluations of alleged maltreatment were interviewed during a 5-day inpatient stay with specific and misleading questions about an anogenital examination and psychological consultation. Support for an age pattern in accuracy was found, showing an increasingly smaller proportion of errors for specific questions from younger to older children (Eisen et al., 2002).
Age has further been found to be a reliable predictor of the amount of information children recall (e.g., Goodman, Quas, Batterman-Faunce, Riddlesburger, & Kuhn, 1994;
Peterson & Whalen, 2001). Goodman et al. (1994) showed the age-effect in a study interviewing 46 children aged 3 to 10 years old within 3 weeks of an invasive medical
procedure (voiding cystourethogram fluoroscopy; VCUG) and found differences in free recall accounts. In the same vein, older children were found to report significantly more
information than younger children at time delays ranging from one-week to 5 years after a traumatic injury occurring at the age of 2 to 13 years requiring treatment at an emergency facility (Peterson & Whalen, 2001). The youngest group typically reported fewer than 50% of details available at all delays, and the oldest typically reporting 80-90% (Peterson & Bell, 1996; Peterson & Whalen, 2001). Former research on maltreated children and disclosure on child abuse, confirm findings on non-maltreated children according to effects of age (Eisen et al., 2002). Empirical findings give support to the first prediction, by which an age effect is expected according to the amount of information reported by the children: Older children will report more information about the removal situation than younger children both shortly after
the separation from their parents and when interviewed three months later. Older children will also report more accurately and show better abilities to reject misleading questions compared to younger children.
Reminders. Traumatic as well as non-traumatic events can be associated with a diverse number of reminders. In the study conducted by Peterson and Whalen (2001) children were likely exposed to reminders at least by accomplishing repeated interviews about the event, providing opportunities for reactivation of the memory. For traumatic events there may be reminders such as media writings (e.g., accidents, natural disasters) and physical evidence if scars, fractions or the like following the trauma (e.g., injuries, medical procedures). In
contrast, highly traumatic experiences are less likely to be talked about. For example in cases of incest, children can be told to keep the abuse a secret or the degree of taboo makes one avoid rather than raise the topic. In their review, Cordón et al. (2004) shed light on the
possibility that reminders might have different effects on implicit versus explicit memories for traumatic events. In implicit memories, as reflected in emotional reactions, behavioral
responses, or preferences, reminders as re-encountering the trauma context, but without the traumatic experience, may attenuate the non-verbal response. In explicit memory on the other hand, reminders may facilitate affective long-term verbal recollection (Cordón et al., 2004).
Research shows that reinstatement and reactivation, along with relatively brief reminders, can be major determinants of whether an experience is forgotten or remains accessible over time (Cordón et al., 2004). Children being removed from home will be
surrounded by reminders of different kinds, some of more script like or semantic quality such as living in a new family and starting in a new kindergarten/school. Other types of reminders will be direct autobiographical, such as talking to and/or visiting their biological parents, getting questions from peers, and professionals about the event to be remembered (e.g., the removal). This is likely to enhance their memory for their situation per se, but it is uncertain whether it concerns their memory for the removal day.
Affecting at least some of the children may be the fact that a lack of knowledge at the time of trauma is found to influence the children’s understanding and appraisal of traumatic events, reported to result in a less durable and detailed event representation (Salmon & Bryant, 2002). According to children being replaced in a foster home or another care facility, adults surrounding them after removal are supposed to be conscious about their experiences and the importance of the separation from their parents resulting in a high degree of dialogue about the experience that may counterbalance the lack of understanding (Goodman & Melinder, 2007a; Nelson & Fivush, 2004).
Degree of stress. Traumatic experiences, especially the core of the event to be remembered, generally appear to be better recalled over longer delays than is typically the case for other experiences (Berntsen, 2002), perhaps because of their distinctiveness or salience (Alexander et al., 2005; Howe et al., 2006b). For highly negative experiences, information directly related to the cause of the stress is prioritized in memory, with such information often retained better the greater the distress (Christianson, 1992). This relation between memory and distress is postulated to imply that increased severity of trauma results in more accurate memory for main features of the event (Alexander et al., 2005). A study with 189 children being interviewed about their traumatic injury and the following hospital
treatment at different time delays, shortly after and numerous times up till 5 years after the incidence, found at least two interesting results according to the estimated stress in the situation and the delay of memories (Peterson & Bell, 1996; Peterson & Whalen, 2001).
When children were interviewed at the 6-month-delay they reported significantly less
information than at the interview shortly after the injury and hospital event (Peterson & Bell, 1996). At later time-delays Peterson and Whalen (2001) found a decrease in memory over time for peripheral information and memories of the treatment at the emergency facility, but a strikingly similar amount of retrieved information of the injury itself between the 6-months delay and 5-year delay.
In the study of young children’s long-term recall of the hurricane Andrew according to how much stress the children were exposed to, Fivush and colleagues (2004) found that the children experiencing moderate degree of stress reported the most when interviewed within a few months after the trauma, but after 6 years all three groups reported more than the first time and they reported the same amount of information. The only difference was that the highly stressed children needed more questions and prompts than children in the other stress groups (Fivush et al., 2004). Overall, childhood traumatic events are found to be subject to forgetting, as are non-traumatic events, but the durability and accessibility of traumatic events may exceed that of neutral and positive events (Cordón et al., 2004), and degree of stress may play an important role.
Experimental studies of adults with PTSD, a disorder characterized by painful recalls, avoidance behavior, and hyperarousal (Copeland, Keeler, Angold, & Costello, 2007), have found a highly accurate long-term memory for the trauma (Alexander et al., 2005). In their prospective study of 94 adult victims of CSA with legal experience resulting, Alexander and colleagues (2005) examined the memory accuracy and errors 12 and 21 years after the abuse ended. The result showed a positive association between accuracy of memory for the trauma
and having severe PTSD symptomatology, also showing better memory for trauma related words and less oblivion for these (Alexander et al., 2005). Although few studies have been conducted with children, it has been found that for maltreated children age 3 to 17 years, PTSD symptoms were associated with more commission errors, but were additionally
associated more correct recall (Eisen et al., 2002). However, others have found no differences in memory performance between children with and without PTSD, leaving an assumption of un-relatedness between PTSD and memory performance in children (Howe et al., 2006a). In this pre-study the relation between having PTSD symptomatology and memory performance is possible to explore, and a logical assumption is to find a positive correlation between PTSD symptoms and the amount of correct information and more commission errors given.
Based on the above reviewed research, children experiencing a high level of stress are expected to remember the same amount of information about the removal day, or more, after three months compared to after one week. According to correctness in what is recalled and reported, children experiencing a high degree of stress are predicted to remember more correct information and less incorrect than those experiencing a low degree of stress.
Not until recently has focus been given to potential effects different kinds of child maltreatment may have on children’s cognitive development (Howe et al., 2006a). So far, factors that have been discussed are the individual factor age and factors characterizing the experience, the traumatic situation.
In addition, there are factors found to have an effect on memory that are also found to be more evident among maltreated than non-maltreated children. For example, psychopathology, lower scores on intelligence measures, lower short-term memory capacity (Eisen et al., 2002), and low socio-economic status (SES) (Howe et al., 2004). Experiencing maltreatment in childhood makes children more
vulnerable for using dissociation as a coping strategy and the chance for dissociation developing to be the dominant strategy used (Cholankeril et al., 2007; van der Kolk et al., 2001). Research on maltreated children, dissociation, and memory have found some association with the factors found to be more evident among maltreated children that will be discussed after looking at challenges in the
conceptualization of dissociation.
The concept of dissociation
The concept of dissociation and its potential protective character has been known since Pierre Janet described women with hysteria in the late 19th century (Janet, 1907/1920), but is still a phenomenon surrounded by controversy and skepticism. Dissociation protects the individual in the face of an overwhelming trauma, and may enable a child to mentally avoid an ongoing trauma he or she cannot physically avoid; dissociation helps the person cope with severe trauma (Cordón et al., 2004; Diseth,
2005). Definitions of dissociation in the literature are diverse, including, “a disruption of usually integrated functions of consciousness, memory, identity, or perception” (APA, 2000, p. 519); a severe deficit in the integration of the self (Macfie, Cicchetti, & Toth, 2001); and, a failure to integrate memories of an event leaving them less accessible to conscious recollection through the formation of isolated memories (Cordón et al., 2004). Central to most definitions of dissociation is a significant change in normal consciousness or awareness that arises from reduced or altered access to one’s
thoughts, feelings, perceptions and/or memories (Briere, Weathers, & Runtz, 2005). Behavior observed in children thought to be reflect dissociation includes the child appearing as withdrawn, like “in a daze”, interpersonally non-responsive, and inattentive. Some are significantly involved in fantasies about themselves and others (Briere, 2005; Macfie et al., 2001). Dissociative defences may be conceptualized as performing three major tasks: automatization of behavior, compartmentalization of painful memories and affects, and estrangement from self, facing potential death or destruction (Putnam, 1997).
Today, two ways of understanding and classifying dissociation exist: a dimensional approach and a categorical approach (Diseth, 2005). Within a dimensional approach, dissociation can be conceptualized as a complex psychological process occurring at a continuum ranging from a wanted kind of minor normative dissociation (e.g., deep concentration, daydreaming), to severe psychopathological conditions (e.g., dissociative identity disorder). Thus, between these two extremes is a continuum where every degree of dissociation may occur (Nijenhuis et al., 2006). This mental strategy becomes pathological when it leads to functional loss or altered behavior (Diseth, 2005). A categorical approach is represented in the classification systems. In ICD-10 dissociation is described in relation to one main category, dissociative (conversion) disorders with eleven subcategories describing mainly altered consciousness (WHO, 1992), not including loss of sensation (e.g., pain), classifying those as somatoform symptoms (Diseth, 2005). DSM-IV-TR includes five
diagnoses under the classification of dissociative disorders: Dissociative amnesia, dissociative fugue, dissociative identity disorder, depersonalization disorder, and dissociative disorder not otherwise specified (APA, 2000). Another perspective described by several authors is the distinction of dissociation in different levels related to the severity of dividedness during the dissociative experience: primary dissociation refers to a dividedness between the normal personal state and the traumatic personal state, a division of the trauma; secondary
dissociation refers to a dividedness within the traumatic state between the observing and the experiencing part of the self; and, tertiary dissociation refers to a dividedness in the person’s identity in the trauma aftermath when adapting to daily life (Benum, 2006).
The specificity of the term dissociation has been affected by problems of both over-
and under-inclusiveness (Nijenhuis et al., 2006). Definitional issues are important to both theory-building and empirical investigations. Delineating and clarifying definitional issues, such as the continuum vs. categorical classification, and whether dissociation is premised to be protective or pathological in nature, is of critical importance for assessment, data
interpretation, and theory building (DePrince & Cromer, 2006).
All children show some degree of dissociation, with younger children being more reliant on this mechanism to cope with stress than older children (Putnam, 1997). Highly dissociative children may be at risk for developing a chronic feeling of depersonalization and derealization, resulting in memories that seem to have a dream-like quality (Cordón et al., 2004). The presence of dissociation in many young children, like having imaginary friends (Putnam, 1997), complicates the distinction between normative and trauma induced
dissociation and complicates the study of dissociation in children.
Child maltreatment, memory, and dissociation
Child maltreatment is found to be a factor in dissociation in preschool-aged children as it is in older children and in adults (Macfie et al., 2001), but controversy exists regarding the impact of maltreatment-related sequel on basic memory processes (see e.g. Eisen et al., 2002; Howe et al., 2006a; Howe et al., 2004). Most studies of maltreatment are conducted on adults reporting on their childhood abuse retrospectively, only a few exists on children identified by the child protective services or other child treatment institutions.
In introducing a caretaker report measure of children’s trauma- and abuse-related symptoms, Briere and colleagues (2001) administered the Trauma Symptom Checklist for Young Children (TSCYC) to a clinical sample of 190 children aged 3 to 12 years old who had at history of maltreatment (Briere, 2001). The children were recruited through child advocacy centers, abuse programs or child trauma centers throughout the United States. The subscale dissociation was found to strongly predict exposure to physical abuse (Briere et al., 2001), indicating that children experiencing physical abuse are particularly prone to relay on dissociation to cope. Other studies have found high dissociation in physically abused preschool-aged children (Macfie et al., 2001), and dissociation associated with any kind of maltreatment but with a distinct difference between the experience of sexual abuse displaying high levels of post-traumatic symptoms and physically abused tending to use dissociation as a primary coping mechanism (Cholankeril et al., 2007). Dissociation subscale is found to predict symptom severity in children with PTSD (Sim et al., 2005). Recent studies have looked at maltreatment, memory and dissociation or PTS symptoms more explicit. What do
these studies tell us?
Attention. One aspect of research on dissociation is the suggested association between divided attention and keeping traumatic memories out of awareness (DePrince & Freyd, 2004). In one study researchers used focused and divided attention memory tasks to assess to what degree 198 low SES abused (with low or high dissociation scores) versus non-abused (with low or high dissociation scores) preschool children aged 4 and 5 years differed in
remembering charged and neutral pictures presented. Results were found to be consistent with the idea that traumatized people may use divided attention to keep threatening information out of awareness (i.e., betrayal trauma theory), data showing that abused children with high dissociation score remembered fewer charged pictures relative to non-abused children (Becker-Blease et al., 2004b). Contrary to the prediction that dissociation is associated with the use of divided attention was the reported similarity between abused children with high and low dissociation score on memory tasks for charged and neutral pictures (Becker-Blease et al., 2004b). The results are consistent with adult studies, in that maltreated children may develop divided attention skills that facilitate coping of stress and trauma. According to these findings, the assumption is that high scores on dissociation lead to less correct information recalled, even if the findings are mixed whether dissociation is a significant factor according to recall (Becker-Blease et al., 2004b).
Dissociation is associated with trauma and often described as a posttraumatic response (e.g., Putnam, 1997). More recent research has started to look at memory processes related to attention that are relevant to PTSD. For example, individual differences in working memory capacity (i.e., the ability to hold and manipulate material in focal attention) appear to be related to the ability to prevent unwanted material from intruding and negatively affecting task performance (Brewin & Holmes, 2003). Brewin and colleagues have in their research found that healthy individuals with greater short-term memory capacity are better at suppressing unwanted thoughts when instructed to do so under experimental conditions, whether these thoughts are neutral or obsessional in nature (cited in Brewin & Holmes, 2003).
These findings may help explain why low intelligence, which is strongly related to working memory capacity, is a risk factor for PTSD (Brewin, Andrews, & Valentine, 2000). In the study conducted by Eisen and colleagues (2002), maltreated children aged 3 to 17 years, were interviewed about a traumatic medical experience and a psychological consultation. The results indicated that both short-term memory (STM) and intellectual ability predicted facets of memory performance. More specifically, STM was significantly associated with memory for the psychological consultation; and, predicted children’s overall memory for the
anogenital examination. In general, children with greater STM spans also performed better on the event memory task (Eisen et al., 2002). In sum, this study found dissociation to be related to better, not worse, memory in children. Based on Eisen et al. (2002) and reasoning on the theoretical implications from PTSD research (Howe et al., 2006a), a positive relation was predicted to emerge between STM capacity and dissociation in the present study.
Multiple risk factors. Multiple risk factors like low SES, disturbed family and social relations, parental psychopathology, and substance abuse are known to be present in many maltreated children’s life (Cholankeril et al, 2007; Howe et al., 2004). In Norway, there have been large descriptive studies conducted earlier on children in the child protective services that are placed in foster care (Havik, 2004).
These studies have reported that maltreated children do not adapt, emotionally and socially, as well as children not being placed in foster care. Long-term exposure to abuse and neglect sets the stage for an increased need of medical, correctional, social and mental health services as the child grows older (van der Kolk, 2005). Research on PTSD generates knowledge about dissociation as a phenomenon and as a post-traumatic response. One study looking at the statistical association between trauma exposure and dissociative symptoms in a normative sample (n = 618) using the Multiscale Dissociation Inventory, revealed significant dissociative symptoms in only 8% of trauma-exposed individuals from the general population (Briere, 2006). But, 90% of those with at least one clinically significant dissociation-scale on the inventory used, reported a trauma history, and significant dissociation was found in only 2% of individuals not reporting a trauma history. Statistics was suggested to indicate that trauma is an important, but insufficient, condition for the development of dissociative symptomatology. Instead, additional risk factors like high posttraumatic stress and/or reduced affect regulation capacities may determine whether trauma exposure results in clinically significant dissociation (Briere, 2006). The relation between memory, dissociation and maltreatment is thus characterized by complexity.
Present study
From about age 3 years on, children are found to give reasonably coherent accounts of past experiences, especially of novel situations (Fivush, 1998). DSM-IV gives a notion about being aware of not
confusing dissociative amnesia with developmentally appropriate childhood amnesia, i.e., the decrease in recall of autobiographical events occurring before the age of 5 (APA, 2000). The children included in this study are removed from home, experiencing being separated from their biological parents. What situations are experienced as traumatic, and the degree of stress experienced, differs among children, but in this study being separated from the primary caregiver and replaced by the CPS is defined as a
traumatic experience.
In Norway, the Government by the child protective services (CPS) is responsible for providing
children who experience lack of sufficient care from their parents’ necessary care at the right time, including temporary orders in acute situations and care orders for an extended period. Children can be placed in a foster home, institution or another suitable day care facility with biological parents´ consent or with force. According to the Child Welfare Act (CWA) an acute removal can be accomplished according to section 4-4, paragraph 5 and section 4-6. A planned removal can be accomplished according to the CWA section 4-4, paragraph 5 or section 4-12. Removal is a severe intervention in a child’s and a family’s life. The main purpose of the CWA is to provide help in the child’s best interest, and it is important to get more knowledge regarding the methods the Government applies in these situations. It is of political, practical and theoretical interest to learn more about the methods and how the children themselves experience, understand and feel about the intervention.
From the introduction outlined above several predictions were advanced. First, age was predicted to correlate positively to the amount and accuracy of information given in the memory interviews. Second, children rated as experiencing a high degree of stress, were predicted to recall the same amount of information, or more, after three months compared to after one week. Third, children experiencing a high degree of stress were predicted to remember more correct information and less incorrect than children experiencing a low degree of stress. Fourth, children experiencing peri-traumatic dissociation and/or with high scores on dissociation and/or PTS total were predicted to recall more than children with low scores. Finally, STM capacity was predicted to correlate positively with degree of
dissociation.
Method Ethical considerations and informed consents
Because this study challenged the established rule for informed consent, a brief reflection and description of the necessary steps is first provided, followed by a presentation of the design and finally the traditional disposition for participants, materials, and procedures.
Ethical considerations. With the permission from the Ministry of Children and
Equality, the Regional Committee for Medical Research Ethics, the Board for Confidentiality and Research, and the Data Inspectorate we thus investigated several issues related to children ages 3-12 years, being removed from their primary caretakers by the CPS. For the planned removals, informed consent was obtained from the biological parents before the actual day of removal according to established rules. But for the acute removals, getting an informed consent before or in the actual situation was problematic. Conducting research in stressful, acute situations demands thorough considerations. The biological parents were often in
conflict with the CPS when an acute removal was accomplished, reacting with overly cooperativeness or withdrawal and hostility. For example, asking for consent via the CPS could have led parents to give their consent as part of their cooperative attitude, regarding the research project to be in the interest of the CPS. For the acute removals a researcher, with experience from similar situations, participated presenting herself as a researcher from the university being there to observe and register what happened. The parents were told that they and/or their lawyer would be contacted for further information. Vulnerable children were involved, and a removal often involves one or more unknown adults to the child in the
situation and at the new residential home. The researcher having a withdrawn neutral position during the removal was essential not to burden the child and family additionally. Ethical considerations have been weighted, and the importance of getting insight into these situations and the knowledge gained were found to be superior, letting a researcher attend without having the parents’ consent.
Informed consent. For the planned removals, informed consent was obtained from the biological parents before the day of removal as described. If one of the parents had parental responsibility but took no part in the care, the relevance of contacting him/her was considered in each case. The caregiver could have another relation to the child, e.g., grandparent, aunt or uncle, but was accounted to be a significant attachment figure for the child. Still the biological parents with parental responsibility for the child would give consent to participation in the project.
For the acute removals, the Ministry of Child and Equality gave the project
responsible Dr. Annika Melinder and two of her staff members, Ragnhild Klingenberg Stokke and Gunn Astrid Baugerud an exception from getting informed consent before or in the removal situation. This allowed the three named researchers to attend the removal and observe what took place, and then contact the biological parents and/or their lawyer shortly after the removal day to get their informed consent. If the parents did not want to participate, the information obtained was maculated.
For each step of the research the participants were informed about the aim and the procedure for the project. The biological parents gave an informed consent according to the general rules (e.g., the Helsinki declaration) and the children gave their assent to participation.
Research involving vulnerable children and adolescents require that the researcher ask the child about his/her opinion (Fisher, 2004). The procedure of consent was a continuous process characterized by an explanation of the research project to the participant in consent-relevant
terms appropriate to the participant’s language preferences (e.g., dialect) and proficiencies, as well as developmental level. This ensured valid consents to an optimum.
Design
The main study, of which the present thesis is part of, is outlined as an experimental study within a cognitive developmental perspective. Specifically, the study composes to a 2
(removal condition; acute removal vs. planned removal) x 3 (interview/observation; one week after the removal vs. three months after vs. one year) mixed factorial design. Control over the experimental situation is obtained through the presence of a researcher during the removals.
This pre-study did not include data from the one-year follow-up. Further, it had a special focus on dissociation as a mechanism in memory for traumatic experiences and limits the report to these main areas of interest. Of course, several other measures were also
collected, such as caregiver attachment style, and the child’s expressive language ability.
Participants
Children and their families were recruited through cooperation with the CPS in three counties in Norway (i.e., Oslo, Akershus, and Buskerud), which include 15 municipalities. The CPS informed the researcher team about current cases of acute removal according to the Child Welfare Act (CWA) §§ 4-4, 5.paragraph, 4-6, 1., or 2.paragraph, and cases of planned removal according to the CWA § 4-4, 5.paragraph or cases being prepared for the count committees for social affairs according to the CWA § 4-12. The main project will include 72 children divided in the acute (n = 36) and planned (n = 36) removal conditions. In the present study, 12 children (n = 12; 5 boys and 7 girls), aged 40-141 months at removal day (M = 89.8, SD = 34.0) are included. At the one-week follow-up data are complete for all participants, but due to one family withdrawing their consent as their child moved back home with the parents and one child being removed for less than three months ago, interview and test data have been collected for 10 children at the three-month follow-up. The participants were recruited from urban and rural areas. The researcher attending the removal can only focus on one child at a time, thus in sibling removals a selection procedure have been employed following this order:
In families with two children within the target group, the youngest/oldest child was included every other time. In families with more than two siblings within the target group, the oldest was included in case 1, the one in the middle in case 2, and the youngest in case 3 (cases being separate removals). If there was an even number of siblings in families where we were going to follow the child in mid position, we estimated the mean age and chose the child
